Regular expression absorbes @CRLFs

[AndyS01]

I have a regular expression that I use in a StringRegExpReplace() function that replaces CRLFs if the matching text is the last text on a line.
Here is my example script:

test()
Exit (1)

Func test()
    Local $sfor, $pat, $sTextBefore = "", $sTextFixed

    $sTextBefore &= "Line 1 MF Midget vs NE Midget" & @CR
    $sTextBefore &= "Line 2 Midget  1 vs VYY Stars" & @CR

    $sfor = "Midget"

    $pat = "(?i)Midget\s{0,}[s]{0,1}\s{0,}[0-9]?+"
    $sTextFixed = StringRegExpReplace($sTextBefore, $pat, "+++")

    ConsoleWrite("+++: $pat ====>" & $pat & "<==" & @CRLF)
    ConsoleWrite("+++: $sTextBefore ==>" & @CRLF & $sTextBefore & @CRLF)
    ConsoleWrite("+++: $sTextFixed ===>" & @CRLF & $sTextFixed & @CRLF)
EndFunc

The output looks like this (I can't figure out how to remove the strikeout in this example):

Quote

+++: $pat ====>(?i)Midget\s{0,}{0,1}\s{0,}[0-9]?+<==
+++: $sTextBefore ==>
Line 1 MF Midget vs NE Midget
Line 2 Midget  1 vs VYY Stars

+++: $sTextFixed ===>
Line 1 MF +++vs NE +++Line 2 +++ vs VYY Stars
 

 

Edited May 28, 2019 by AndyS01

[Jos]

Is this what you want? :

$pat = "(?i)Midget\s{0,}?[s]{0,1}?\s{0,}?[0-9]?+"

Jos

[AndyS01]

Almost.  Adding '?' after a match prevents regexp from being greedy?  The concept of greedy, lazy and possessive is quite confusing.  I looked at some on line tutorials and got my head swimming.

My example didn't supply a plural for Midget.  I meed to match 'Midget', 'Midgets', 'Midget 1', 'Midgets 1':

   $sTextBefore &= "Line 1 MF Midgets vs NE Midgets" & @CR
    $sTextBefore &= "Line 2 Midgets 1 vs VYY Stars" & @CR

When I do that, I only match 'Midget', not 'Midgets', like so:

test()
Exit (1)

Func test()
    Local $sfor, $pat, $sTextBefore = "", $sTextFixed

    $sTextBefore &= "Line 1 MF Midgets vs NE Midgets" & @CR
    $sTextBefore &= "Line 2 Midgets  1 vs VYY Stars" & @CR

    $sfor = "Midget"

    $pat = "(?i)Midget\s{0,}?[s]{0,1}?\s{0,}?[0-9]?+"
    $sTextFixed = StringRegExpReplace($sTextBefore, $pat, "+++")

    ConsoleWrite("+++: $pat ====>" & $pat & "<==" & @CRLF)
    ConsoleWrite("+++: $sTextBefore ==>" & @CRLF & $sTextBefore & @CRLF)
    ConsoleWrite("+++: $sTextFixed ===>" & @CRLF & $sTextFixed & @CRLF)
EndFunc

The output is this (notice the 's' characters remaining in the output):
 

Quote

Line 1 MF +++s vs NE +++s
Line 2 +++s  1 vs VYY Stars

 

[Jos]

It's always fun these regex's

What about this version:

$pat = "(?i)(?m)Midgets*[[:blank:]]+[0-9]*[[:blank:]]*"

Jos

[AndyS01]

Perfect, just what I needed.  Thank you.

[iamtheky]

even though it doesnt catch the second instance of midget in Line 1?

Quote

+++: $pat ====>(?i)(?m)Midgets*[[:blank:]]+[0-9]*[[:blank:]]*<==

+++: $sTextBefore ==>
Line 1 MF Midget vs NE Midget Line 2 Midget  1 vs VYY Stars

+++: $sTextFixed ===>
Line 1 MF +++vs NE Midget Line 2 +++vs VYY Stars

 

Here's one minus the regex, just hammers the pieces into place

test()
Exit (1)

Func test()
    Local $sfor, $pat, $sTextBefore = "", $sTextFixed

    $sTextBefore &= "Line 1 MF Midget vs NE Midget" & @CR
    $sTextBefore &= "Line 2 Midget  1 vs VYY Stars" & @CR

    consolewrite(StringReplace(StringStripCR(StringReplace($sTextBefore , "Midgets" , "+++")) , "Midget" , "+++"))

EndFunc

 

Edited May 28, 2019 by iamtheky

[AndyS01]

Oh-Oh  I see what you mean iamtheky. This didn't work:

test()
Exit (1)

Func test()
    Local $sfor, $pat, $sTextBefore = "", $sTextFixed

    $sTextBefore &= "Line 1 MF Midget vs NE Midgets" & @CR
    $sTextBefore &= "Line 2 Midgets  1 vs VYY Stars" & @CR

    $sfor = "Midget"

    $pat = "(?i)(?m)Midgets*[[:blank:]]+[0-9]*[[:blank:]]*"
    $sTextFixed = StringRegExpReplace($sTextBefore, $pat, "+++")

    ConsoleWrite("+++: $pat ====>" & $pat & "<==" & @CRLF)
    ConsoleWrite("+++: $sTextBefore ==>" & @CRLF & $sTextBefore & @CRLF)
    ConsoleWrite("+++: $sTextFixed ===>" & @CRLF & $sTextFixed & @CRLF)
EndFunc

The output was:

Quote

+++: $pat ====>(?i)(?m)Midgets*[[:blank:]]+[0-9]*[[:blank:]]*<==
+++: $sTextBefore ==>
Line 1 MF Midget vs NE Midgets
Line 2 Midgets  1 vs VYY Stars

+++: $sTextFixed ===>
Line 1 MF +++vs NE Midgets
Line 2 +++vs VYY Stars

 

[AndyS01]

The 's' in Midgets' is optional.  As I said before, I want to replace "Midget", "Midgets", "Midget 1", "Midgets 1" with the "+++" string.  Also, the '1' could be [0-9].

[jchd]

Then that should work for you:

_test()
Exit (1)

Func _test()
    Local $sfor, $pat, $sTextBefore = "", $sTextFixed

    $sTextBefore &= "Line 1 MF Midget vs NE Midgets" & @CR
    $sTextBefore &= "Line 2 Midgets  1 vs VYY Stars" & @CR

    $sfor = "Midget"

    $pat = "(?im)Midgets?(?: *\d*)?"
    $sTextFixed = StringRegExpReplace($sTextBefore, $pat, "+++")

    ConsoleWrite("+++: $pat ====>" & $pat & "<==" & @CRLF)
    ConsoleWrite("+++: $sTextBefore ==>" & @CRLF & $sTextBefore & @CRLF)
    ConsoleWrite("+++: $sTextFixed ===>" & @CRLF & $sTextFixed & @CRLF)
EndFunc

 

[mikell]

An alternative way allowing variations

$sTextBefore = "Line 1 MF Midget vs NE Midgets" & @CR & _
        "Line 2 Midgets -1a vs VYY Stars" & @CR

$pat = "(?im)Midget.*?(?=\hvs|$)"
Msgbox(0,"", StringRegExpReplace($sTextBefore, $pat, "+++"))

 

[AndyS01]

2 hours ago, mikell said:

An alternative way allowing variations

$sTextBefore = "Line 1 MF Midget vs NE Midgets" & @CR & _
        "Line 2 Midgets -1a vs VYY Stars" & @CR

$pat = "(?im)Midget.*?(?=\hvs|$)"
Msgbox(0,"", StringRegExpReplace($sTextBefore, $pat, "+++"))

 

In real live production, the 'vs' might not be there.

 

2 hours ago, mikell said:

An alternative way allowing variations

$sTextBefore = "Line 1 MF Midget vs NE Midgets" & @CR & _
        "Line 2 Midgets -1a vs VYY Stars" & @CR

$pat = "(?im)Midget.*?(?=\hvs|$)"
Msgbox(0,"", StringRegExpReplace($sTextBefore, $pat, "+++"))

 

 

[AndyS01]

The example in jchd's reply worked for all examples that I tried.  Thanks

[mikell]

5 hours ago, AndyS01 said:

In real live production, the 'vs' might not be there.

So the question was badly asked. No way to guess that "real life" might be different than the provided example 

[AndyS01]

True enough.  The original question was why the CRLF was being absorbed, joining two lines,  The eventual fix was the (?m) part of the RE.